A TUMOR NECROSIS FACTOR-BETA GENE POLYMORPHISM IN RELATION TO MONOKINE SECRETION AND INSULIN-DEPENDENT DIABETES-MELLITUS

HLA-class III region genes may be associated with susceptibility to insulin-dependent diabetes mellitus (IDDM). In this study an NcoI polymorphism of the tumour necrosis factor beta (TNF-beta) gene, which is positioned next to the tumour necrosis factor alpha (TNF-alpha) gene in the HLA class III region, was detected by restriction fragment length polymorphism (RFLP). This polymorphism has previously been reported to be located in the TNF-alpha-gene. Caucasian HLA-DR3,4 heterozygous IDDM patients (n = 26) and DR-matched healthy controls (n = 19), as well as randomly selected IDDM patients (n = 27) and controls (n = 25) were studied. In addition four multiplex families (49 individuals) and eight HLA-non-identical sibpairs concordant for IDDM were analysed. The TNF-beta-gene RFLP analysis showed fragments of 5.5 kb and 10.5 kb, which behaved as alleles. In all groups there was a haplotype assignment of the TNF-beta-5.5-kb allele to B8,DR3 haplotypes, and of the TNF-beta-10.5-kb allele to B15,DR4-positive haplotypes. The allelic and genotypic frequencies differed between DR3,4 IDDM patients and DR3,4 controls, and the DR3,4 control group differed significantly from the randomly selected control group (P < 0.0079). In HLA-DR3,4- and DQw8-positive persons, the DR3 haplotypes carried the 10.5-kb allele three times more frequently in IDDM patients than in controls, suggesting that the 10.5-kb allele when present on DR3 haplotyes may contribute to susceptibility to IDDM in DR3,4 heterozygous individuals. A contributory role of the 10.5-kb allele in genetic IDDM susceptibility was supported by the sibpair analysis, in which all were TNF-beta-identical. Fiver were 10.5 kb homozygous, and the remaining three pairs were 5.5/10.5 kb heterozygous. Twenty-five healthy and eight newly diagnosed IDDM patients were randomly selected to study the Escherichia coli lipopolysaccharides (LPS)-purified protein derivative (tuberculin) (PPD)-, and phytohaemagglutinin (PHA)-stimulated monocyte (Mo) secretions of interleukin 1 beta (IL-1-beta) and TNF-alpha in relation to the NcoI TNF-beta-gene polymorphism. The LPS- and PHA-stimulated Mo IL-1-beta and TNF-alpha-secretions were significantly lower for the TNF-beta-5.5/10.5 kb heterozygous individuals than for TNF-beta-10.5 kb homozygous individuals. Furthermore, the Mo IL-1-beta and TNF-alpha-secretions of IDDM patients were significantly higher than the Mo secretions of TNF-beta-genotype-matched healthy controls. This study suggests an association between the 10.5 kb TNF-beta-allele and IDDM, and demonstrates an association between monokine responses and TNF-beta-genotypes. These observations may have implications for understanding the pathogenesis of HLA-associated autoimmune diseases including IDDM.